This article explores the use of machine learning methods in vibrodiagnostics. To assess the effectiveness of the use of a neural network, a hypothesis was put forward on the possibility of revealing hidden patterns indicating a defect for structures of the same type with different parameters. The selection of these features by visual analysis of the graphs is difficult due to the large amount of data, which indicates the relevance of solving the classification problem using a neural network. As a result of the study, a fully-connected two-layer neural network was obtained, showing sufficient accuracy of predictions, which confirms the fundamental possibility of using machine learning methods to monitor the state of standard structures.

Keywords: vibration diagnostics, building construction, building, structure, defect, damage, machine learning, neural network

The question of the diagnosis of the state of building structures using non-destructive testing. These methods allow to predict the condition of structures to prevent accidents and, therefore, increase the service life of structures. The main objective of the study - assessment of changes in the vibration mode shapes method for the detection of structural damage , in the absence of forced uploading. The article describes the implementation of a mathematical method of vibration identification and localization of lesions of metal trusses . The results of the calculation for fault location triangular steel truss.

Keywords: Vibration diagnostics, fault detection, diagnosis, construction construction spans, trusses, trusses, defect, damage, vibration, localization, identification

The last ten years have seen a rapid growth of research related to the development of methods of localization of defects and assessment of technical condition of metal structures based on dynamic analysis. This paper presents the results of a study on the localization of defects arched steel truss. The main purpose of this article is to analyze the results of the damage index method and the development of the software module to localize structural damage. The results of numerical experiments presented in this article have shown the effectiveness of this method of localization of damage. The use of this method in practice will reduce the time of inspection by pre- defining a site design that are necessary to pay special attention during the field study.

Keywords: Vibration diagnostics, fault detection, diagnosis, construction construction spans, trusses, trusses, defect, damage, vibration, localization, identification

The possibilities of the vibration method changes the curve oscillation localization damage concrete structural elements with prestressed reinforcement. This article discusses a method for curve oscillation example concrete slab in terms of spatial, complex stress state under the action of vertical force Q.

Keywords: Vibration diagnostics, fault detection, diagnosis, design, construction span, reinforced concrete, concrete, reinforcement, prestressing.

`

Keywords: